INCA Sensor: Soft sensors for non-measurable quality variables

Online determination of quality variables

Production plants in the process industry today rely on regular and extremely time-consuming laboratory analyses for quality control purposes (new measured values typically every 8 to 24 hours). Or they use very expensive and high-maintenance online analyzers (new values typically every 20 to 60 minutes). In order to raise productivity and run the process up to its full capacity while maintaining the required quality, it is necessary to measure product quality online with a refresh rate of between 0.5 and 3 minutes. This ensures that the controller responds at the right time and the product specifications are maintained.

The plant operator is usually responsible for complete or partial process control. Weak points in process control are also reflected by changes in product quality or in the utilization of production capacity. Since no quality values are known during a conversion, the production specifications do not correspond to the quality requirements over longer periods.

These problems can be solved by using soft sensors. Soft sensors are calculation procedures which determine non-measurable quality variables on the basis of measurable process values (pressures, flow rates, temperatures, levels, etc.) in cycles of between 0.5 and 3 minutes. The calculation is made on the basis of a (non-)linear parametric model generated from historic plant data or through dedicated tests. The high-speed soft sensor predictions can be consolidated by laboratory analyses or values from online analyzers.

The soft sensor predictions enable the frequency of laboratory analyses and the use of online analyzers to be reduced. They raise product quality while at the same time reducing operating costs.

INCA Sensor is a tool for designing, parameterizing and operating soft sensors. It makes it easier to master complex plant dynamics, and enables operating conditions to be optimized so that the quality of the end product is assured.

Note:

INCA Sensor can be used in combination with SIMATIC PCS 7 V8.X and V9.0. As of SIMATIC PCS 7 V8.2, INCA MPC or INCA Sensor can be integrated via the Advanced Process Control (APC) coupling modules of the SIMATIC PCS 7 Industry Library (as of V8.2 Update 1).

Application examples

• Polymer thickness
• Polymer melt-flow index
• Viscosity
• Product concentration at the outlet of reaction or distillation columns
• Plant efficiency/utilization factor
• Cement properties
• Exhaust parameters of combustion processes
• Gas concentrations (NO_x, CO₂, etc.)

INCA Sensor sets new standards for the permanent plant-wide optimization and control of non-linear processes. INCA Sensor differs from other soft-sensor program packages due to its series of function expansions that support the designer when drafting reliable soft sensors:

• Modern modeling methods such as linear transmission functions, general non-linear models (GNOMOs) or estimates according to the partial least squares estimators method
• Signal processing or pre-processing (offline and online)
• Powerful tools for selecting suitable input variables
• Input options for data from laboratories and analyzers

Soft sensors are a prerequisite for plant optimization and quality control using advanced process control solutions such as INCA MPC.

The INCA Sensor program package can run on a separate PC/server with Windows 7/8/10 or Windows Server 2012/2016 operating systems. It is linked to the SIMATIC PCS 7 process control system by means of OPC, where INCA Sensor is operated as an OPC client.

INCATOOLS

Phone: +32 1639 3083
Fax: +32 1639 3080

Email: christiaan.moons@incatools.com

You can find more information on the Internet at:
http://www.incatools.com